Plot.VariableImportance {AutoPlots}

Plot.VariableImportance

Description

Generate variable importance plots

Usage

Plot.VariableImportance(
  dt = NULL,
  XVar = NULL,
  YVar = NULL,
  GroupVar = NULL,
  YVarTrans = "Identity",
  XVarTrans = "Identity",
  FacetRows = 1,
  FacetCols = 1,
  FacetLevels = NULL,
  AggMethod = "mean",
  Height = NULL,
  Width = NULL,
  Title = "Variable Importance Plot",
  ShowLabels = FALSE,
  Title.YAxis = NULL,
  Title.XAxis = NULL,
  EchartsTheme = "macarons",
  TimeLine = TRUE,
  TextColor = "white",
  title.fontSize = 22,
  title.fontWeight = "bold",
  title.textShadowColor = "#63aeff",
  title.textShadowBlur = 3,
  title.textShadowOffsetY = 1,
  title.textShadowOffsetX = -1,
  xaxis.fontSize = 14,
  yaxis.fontSize = 14,
  Debug = FALSE
)

Arguments

dt

source data.table

XVar

Column name of X-Axis variable. If NULL then ignored

YVar

Column name of Y-Axis variable. If NULL then ignored

GroupVar

Column name of Group Variable for distinct colored histograms by group levels

YVarTrans

"Asinh", "Log", "LogPlus1", "Sqrt", "Asin", "Logit", "PercRank", "Standardize", "BoxCox", "YeoJohnson"

XVarTrans

"Asinh", "Log", "LogPlus1", "Sqrt", "Asin", "Logit", "PercRank", "Standardize", "BoxCox", "YeoJohnson"

FacetRows

Defaults to 1 which causes no faceting to occur vertically. Otherwise, supply a numeric value for the number of output grid rows

FacetCols

Defaults to 1 which causes no faceting to occur horizontally. Otherwise, supply a numeric value for the number of output grid columns

FacetLevels

Faceting rows x columns is the max number of levels allowed in a grid. If your GroupVar has more you can supply the levels to display.

AggMethod

Choose from 'mean', 'sum', 'sd', and 'median'

Height

"400px"

Width

"200px"

Title

title

ShowLabels

character

Title.YAxis

character

Title.XAxis

character

EchartsTheme

"auritus","azul","bee-inspired","blue","caravan","carp","chalk","cool","dark-bold","dark","eduardo", #' "essos","forest","fresh-cut","fruit","gray","green","halloween","helianthus","infographic","inspired", #' "jazz","london","dark","macarons","macarons2","mint","purple-passion","red-velvet","red","roma","royal", #' "sakura","shine","tech-blue","vintage","walden","wef","weforum","westeros","wonderland"

TimeLine

logical

TextColor

'darkblue'

title.fontSize

22

title.fontWeight

"bold"

title.textShadowColor

'#63aeff'

title.textShadowBlur

3

title.textShadowOffsetY

1

title.textShadowOffsetX

-1

xaxis.fontSize

14

yaxis.fontSize

14

Debug

Debugging purposes

Value

plot

Author(s)

Adrian Antico

See Also

Other Model Evaluation: Plot.BinaryMetrics(), Plot.Calibration.Box(), Plot.Calibration.Line(), Plot.ConfusionMatrix(), Plot.Gains(), Plot.Lift(), Plot.PartialDependence.Box(), Plot.PartialDependence.HeatMap(), Plot.PartialDependence.Line(), Plot.ROC(), Plot.Residuals.Histogram(), Plot.Residuals.Scatter(), Plot.ShapImportance()